This study focused on the pathways of CA-Bacteriuria. The exoluminal pathway was predominant and accounted for more than one third of cases of early exoluminal CA-Bacteriuria. This result was not related to the quality of urinary catheterization.
CA-Bacteriuria occurred in one third of patients of the Medical ICU, a proportion consistent with that in other ICU studies in [23, 24]. We observed a greater number of cases of exoluminal CA-Bacteriuria (88.5%) and a low rate of endoluminal CA-Bacteriuria (11.5%). In our study, the main pathway was exoluminal and occurred at a higher rate than in the study of Tambyah et al., in which 46%, 23.2% and 30.8% of CA-Bacteriuria cases were exoluminal, endoluminal and indeterminate CA-Bacteriuria, respectively [9]. The authors reported a 10% rate of disconnections between urinary catheters and collector bags, which is a risk factor for endoluminal CA-Bacteriuria. During the four months of our study, only two disconnections (3%) were observed, which could explain the very low proportion of endoluminal CA-Bacteriuria cases. These findings suggest that the guidelines concentrate attention on preventing disconnection thereby leading to a large reduction in the endoluminal pathway [1, 2]. Another explanation could be the increasing use of 100% silicon urinary catheters, which delays the obstruction caused by biofilm encrustation [27].
Tambyah et al. observed 12.4% (n = 31) of early exoluminal catheter-associated bacteriuria among 250 identified microorganisms [9] as against 38.5% in our study. Early exoluminal colonization could be related to a defect in the aseptic procedure during insertion of the catheter. To assess this assumption, we performed a quality audit on catheter insertion in accordance with the most recent guidelines [2, 4, 19]. The only flaw observed was a low observance of hydro-alcoholic hand rubbing (approximately 50%) during the different stages of insertion where alcohol-based products are recommended [26]. This failure was not a major infringement of the guidelines but could have been involved in early exoluminal CA-Bacteriuria. However, the microorganisms of early exoluminal colonization in our study did not come only from the skin flora. Another cause of early exoluminal CA-Bacteriuria could be the presence of microorganisms in the final centimeters of the urethra carried away during catheterization. In our study, 69.5% of the micro-organisms of exoluminal CA-Bacteriuria were also identified on samples of urinary meatus. This is consistent with other reports which showed that in 75% of cases the microorganisms of exoluminal bacteriuria preexisted in the urethral flora [15, 16] and that a positive culture of urinary meatus increased the incidence of CA-Bacteriuria [17, 18, 27]. However, meatal care does not result in a reduction in CA-UTI or urethral colonization [1, 20]. Before insertion, there is a physical inability to access the inside of the urethra during the antisepsis [9]
After insertion, the persistence of CA-Bacteriuria could be related to the formation of biofilm at the interface of the catheter and the urethra, to repeated catheter manipulations or to inadequate residual activity of the antiseptic [1]. Consequently, the strict implementation of guidelines by health workers does not avoid early exoluminal CA-Bacteriuria.
The first limitation of our study is the low number of patients included. However, this was compensated by the large number of samples (n = 1008 with a maximum of 27 samples for 15 days), which is indicative of a good follow-up of each patient. A second limitation was the viable non-cultivable bacteria that we were unable to isolate. To reduce this bias, it would have been necessary to remove the urinary catheter and culture its end or to carry out a molecular analysis by 16S PCR. However, as we were performing a real-life observational study we used the usual microbial analysis techniques. However, this is the first report to confirm microbial results by molecular comparison of the same microorganisms isolated. A final limitation was the choice of a threshold of 102 CFU/mL in bladder urine to define CA-Bacteriuria: in French and American guidelines, it is set at 103 CFU/ml [1, 4]. This could have artificially increased the detection of CA-Bacteriuria. However, our aim was earlier detection of bacteriuria. The change in threshold was possible because standardization of the microbiological technique allowed a detection threshold of 10 CFU/mL. We do not rule out an impact of anti-infective therapy on the detection threshold. However, there was no difference of anti-infective therapy between the two groups. In addition, it is established that in patients with a urinary catheter not receiving antimicrobial therapy bacteriuria or candiduria ≥ 102 CFU/ml will increase to > 105 CFU/mL in one to three days if the urinary catheter remains in place [1, 4].